Over 20 million women in the United States suffer from urinary incontinence, a condition leading to depression, social isolation, functional decline, and loss of ability to live independently. First-line treatment for incontinence consists of behavioral management strategies such as pelvic floor muscle exercises that are difficult for women to lean to practice effectively in course of routine care. Second-line treatments have problematic side effects or other limitations that decrease their safety and accessibility for many women. As a result, alternate treatment strategies are needed that are not only effective but also more accessible and better tolerated. Yoga is a complementary behavioral intervention with the potential to improve bladder control through multiple mechanisms. When practiced in a way that emphasizes mindful awareness of specific bodily structures, yoga can be used to increase women's control over their pelvic floor muscles, improve their autonomic nervous system balance, and decrease underlying stress and anxiety, all of which may alleviate incontinence. To explore the preliminary efficacy and safety of yoga to treat incontinence, our research team developed a structured, group-based yoga therapy program involving twice weekly group yoga classes supplemented by once weekly home yoga practice. In a pilot trial of 19 ambulatory women aged 46 to 75 years with frequent stress, urgency, or mixed incontinence, women randomized to this program reported a average 66% decrease in the frequency of incontinence over 6 weeks, versus a 16% decrease among those randomized to control (P=.05). All women starting the yoga program attended more than 90% of group classes, and all but one reported that it would be easy to continue practicing yoga to treat their urine leakage. We now propose to conduct a 12-week randomized trial to provide rigorous evidence of whether yoga training can effectively and safely enhance usual first-line behavioral management for incontinence. All women will receive written information and brief in-person instruction on patient-directed pelvic floor muscle exercises, bladder training, and other behavioral techniques consistent with usual first-line care. Women randomized to yoga therapy will also take part in twice-weekly group classes and once weekly home practice sessions focused on selected yoga postures and techniques, while those randomized to control will engage in equivalent-time stretching classes and home stretching sessions that will control for participants' time and attention. In addition to examining changes in incontinence frequency, bothersomeness, and impact, we will examine whether changes in incontinence are mediated by changes in pelvic floor strength, autonomic balance, or stress and anxiety. If successful, this research may reduce the burden of one of the most common chronic conditions in women, as well as provide new insights into mechanisms underlying treatment response.